Finite state machine for vehicle detection in highway surveillance systems

In this paper, a novel vehicle detection approach based on a finite state machine is presented. Vehicle detection is an essential function to build road monitoring and traffic surveillance systems. Based on an analysis of a selected road region, vehicle detection problem is attacked by constructing a finite state machine. The proposed system starts by selecting an AOI and checking statistics inside of the AOI. In the proposed system, the ratio of the foreground mask or the number of lighting pixels in the AOI is calculated and fed into the proposed finite state machine. Simulation results show that the proposed system is simple and robust and can be applicable for highway surveillance systems.